This area covers all topics related to the remote sensing and GNSS technologies. All topics are useful for Earth Observation research and for monitoring and mapping the planet or parts of it. GNSS technology instead takes care of positioning both for monitoring and navigation aspects. GNSS signals can be suitable not only for positioning but also, using signal crossing the atmosphere continuously to perform analysis for the atmospheric studies (both of the upper layer or the lowest one).
The image analysis regards the output of a computer vision system, relying on sensors’ data fusion with visual information. The long-term expertise in preprocessing techniques and quantitative imaging permits to denoise and “clean” images to extract quantitative information, using vision sensors as measurement tools. For instance, the tracking techniques allow performing several applications such as reliable and accurate high speed motion detection and analysis, self-localization, attitude determination, pose estimation, trajectory reconstruction and accurate 2D/3D mosaicking. Capable to work under severe light changes. The expertise in virtual, augmented and mixed reality can be also fruitfully employed to build planet surface simulators.
The computer vision group (CVG) can design multi-camera and multi-sensor computer vision systems based on 2D or 3D (stereoscopy, time-of flight, RGBD, lidar, …) imaging, for real-time or batch image acquisition, also utilizing GPU boards or low power devices for embedded systems. The systems are designed to interact with several kinds of sensors (IMU, thermal, ....) exploiting different sensing technology (visible, IR, thermal, ToF) to provide a complete picture of the imaged scene. Experienced in working with on-board cameras.
Pattern recognition and machine learning are applied to the output of the features extracted during image analysis to provide scene comprehension and understanding. Several artificial intelligence techniques were used for supervised and unsupervised classification aiming at discovery interesting properties regarding the objects of interest or, in general, the scene. These techniques are also used for image processing and analysis to perform smart denoising, image enhancement, and so on. Experienced with reliable statistical analysis and classifications using small samples.
Develoment of data processing methods for monitoring application of land and atmosphere. Research on using GNSS topographical survey but also for land, off-shore and space navigation applications in real time.
Techniques and methods for image processing in the optical, hyperspectral or radar fields for mapping the territory and its modifications using advanced methods also based on deep learning and computer vision techniques.
Scientific responsible of the GNSS and Earth Observation operative unit
To receive information about the services of this unit, inquiry at the following email